Analyzing Argumentative Discourse Units in Online Interactions Workshop

This workshop delves into the analysis of argumentative discourse units in online interactions, emphasizing the segmentation, classification, and relation identification processes. It discusses challenges in annotation and proposes a two-tiered annotation scheme involving expert annotators and novices. The approach integrates coarse-grained expert annotation with pragmatic argumentation theory to enhance annotation quality.

• Online Interactions
• Argumentative Discourse
• Annotation Scheme
• Expert Annotators
• Argumentation Mining

horsfall_e Follow

Uploaded on Sep 22, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. Analyzing Argumentative Discourse Units in Online Interactions Debanjan Ghosh, Smaranda Muresan, Nina Wacholder, Mark Aakhus and Matthew Mitsui First Workshop on Argumentation Mining, ACL June 26, 2014

2. But when we first tried the iPhone it felt natural immediately, we didn't have to 'unlearn' old habits from our antiquated Nokias & Blackberrys. That happened because the iPhone is a truly great design. when we first tried the iPhone it felt natural immediately, User1 That's very true. With the iPhone, the sweet goodness part of the UI is immediately apparent. After a minute or two, you re feeling empowered and comfortable. Feeling empowered and comfortable. Feeling empowered and comfortable. That s very true. With the iPhone, the sweet goodness part of That s very true. With the iPhone, the sweet goodness part of The UI is immediately apparent. After a minute or two, you re The UI is immediately apparent. After a minute or two, you re User2 It's the weaknesses that take several days or weeks for you to really understanding and get frustrated by. I disagree that the iPhone just "felt natural immediately"... In my opinion it feels restrictive and over simplified, sometimes to the point of frustration. Point of frustration. Point of frustration. I disagree that the iPhone just felt natural immediately in my I disagree that the iPhone just felt natural immediately in my Opinion it feels restrictive and over simplified, sometimes to the Opinion it feels restrictive and over simplified, sometimes to the User3 Argumentative Discourse Units (ADU; Peldszus and Stede, 2013) 1. Segmentation 2. Segment Classification 3. Relation Identification

3. Annotation Challenges A complex annotation scheme seems infeasible The problem of high *cognitive load* (annotators have to read all the threads) High complexity demands two or more annotators Use of expert annotators for all tasks is costly 3

4. Our Approach: Two-tiered Annotation Scheme Coarse-grained annotation Expert annotators (EAs) Annotate entire thread Fine-grained annotation Novice annotators (Turkers) Annotate only text labeled by EAs 4

5. Our Approach: Two-tiered Annotation Scheme Coarse-grained annotation Expert annotators (EAs) Annotate entire thread Fine-grained annotation Novice annotators (Turkers) Annotate only text labeled by EAs 5

6. Coarse-grained Expert Annotation Target Post1 Post2 Post2 Post3 Post3 Post4 Callout Pragmatic Argumentation Theory (PAT; Van Eemeren et al., 1993) based annotation 6

7. ADUs: Callout and Target A Calloutis a subsequent action that selects all or some part of a prior action (i.e., Target) and comments on it in some way. A Targetis a part of a prior action that has been called out by a subsequent action. 7

8. Target But when we first tried the iPhone it felt natural immediately, we didn't have to 'unlearn' old habits from our antiquated Nokias & Blackberrys. That happened because the iPhone is a truly great design. when we first tried the iPhone it felt natural immediately, User1 That's very true. With the iPhone, the sweet goodness part of the UI is immediately apparent. After a minute or two, you re feeling empowered and comfortable. Feeling empowered and comfortable. Feeling empowered and comfortable. That s very true. With the iPhone, the sweet goodness part of That s very true. With the iPhone, the sweet goodness part of The UI is immediately apparent. After a minute or two, you re The UI is immediately apparent. After a minute or two, you re User2 Callout It's the weaknesses that take several days or weeks for you to really understanding and get frustrated by. I disagree that the iPhone just "felt natural immediately"... In my opinion it feels restrictive and over simplified, sometimes to the point of frustration. Point of frustration. Point of frustration. I disagree that the iPhone just felt natural immediately in my I disagree that the iPhone just felt natural immediately in my Opinion it feels restrictive and over simplified, sometimes to the Opinion it feels restrictive and over simplified, sometimes to the User3 Callout

9. More on Expert Annotations and Corpus Five Annotators were free to choose any text segment to represent an ADU Four blogs and their first one-hundred comment sections are used as our argumentative corpus Android (iPhone vs. Android phones) iPad (usability of iPad as a tablet) Twitter (use of Twitter as a micro-blog platform) Job Layoffs (layoffs and outsourcing) 9

10. Inter Annotator Agreement (IAA) for Expert Annotations P/R/F1 based IAA (Wiebe et al., 2005) exact match (EM) overlap match (OM) a Krippendorff s (Krippendorff, 2004) a Thread F1_EM F1_OM Krippendorff s Android 54.4 87.8 0.64 iPad 51.2 86.0 0.73 Layoffs 51.9 87.5 0.87 Twitter 53.8 88.5 0.82 10

11. Issues Different IAA metrics have different outcome It is difficult to infer from IAA that what segments of the text are easier or harder to annotate 11

12. Our solution: Hierarchical Clustering We utilize a hierarchical clustering technique to cluster ADUs that are variant of a same Callout # of Expert Annotator/ADUs per cluster 5 4 3 # of Clusters Thread 2 1 Android 91 52 16 11 7 5 Ipad 88 41 17 7 13 10 Layoffs 86 41 18 11 6 10 Twitter 84 44 17 14 4 5 Clusters with 5 and 4 annotators shows Callouts that are plausibly easier to identify Clusters selected by only one or two annotators are harder to identify 12

13. Example of a Callout Cluster 13

14. Motivation for a finer-grained annotation What is the nature of the relation between a Callout and a Target? Can we identify finer-grained ADUs in a Callout? 14

15. Our Approach: Two-tiered Annotation Scheme Coarse-grained annotation Expert annotators (EAs) Annotate entire thread Fine-grained annotation Novice annotators (Turkers) Annotate only text labeled by EAs 15

16. Novice Annotation: task 1 T T CO CO Agree/Disagree/Other T T CO CO This is related to annotation of agreement/disagreement (Misra and Walker, 2013; Andreas et al., 2012) identification research. 16

17. Target But when we first tried the iPhone it felt natural immediately, we didn't have to 'unlearn' old habits from our antiquated Nokias & Blackberrys. That happened because the iPhone is a truly great design. when we first tried the iPhone it felt natural immediately, User1 That's very true. With the iPhone, the sweet goodness part of the UI is immediately apparent. After a minute or two, you re feeling empowered and comfortable. Feeling empowered and comfortable. Feeling empowered and comfortable. That s very true. With the iPhone, the sweet goodness part of That s very true. With the iPhone, the sweet goodness part of The UI is immediately apparent. After a minute or two, you re The UI is immediately apparent. After a minute or two, you re User2 Callout It's the weaknesses that take several days or weeks for you to really understanding and get frustrated by. I disagree that the iPhone just "felt natural immediately"... In my opinion it feels restrictive and over simplified, sometimes to the point of frustration. Point of frustration. Point of frustration. I disagree that the iPhone just felt natural immediately in my I disagree that the iPhone just felt natural immediately in my Opinion it feels restrictive and over simplified, sometimes to the Opinion it feels restrictive and over simplified, sometimes to the User3 Callout

18. More from Agree/Disagree Relation Label For each Target/Callout pair we employed five Turkers Fleiss Kappa shows moderate agreement between the Turkers 143 Agree/153 Disagree/50 Other data instance We run preliminary experiments for predicting the relation label (rule based, BoW, Lexical Features ) Best results (F1): 66.9% (Agree) 62.9% (Disagree) 18

19. Novice Annotation: task 2 T S R CO 2: Identifying Stance vs. Rationale Difficulty This is related to identification of justification task (Biran and Rambow, 2011) 19

20. That's very true. With the iPhone, the sweet goodness part of the UI is immediately apparent. After a minute or two, you re feeling empowered and comfortable. Feeling empowered and comfortable. That s very true. With the iPhone, the sweet goodness part of The UI is immediately apparent. After a minute or two, you re That s very true User2 It's the weaknesses that take several days or weeks for you to really understanding and get frustrated by. I disagree that the iPhone just "felt natural immediately"... In my opinion it feels restrictive and over simplified, sometimes to the point of frustration. Point of frustration. I disagree that the iPhone just felt natural immediately I disagree that the iPhone just felt natural immediately in my Opinion it feels restrictive and over simplified, sometimes to the User3 Stance Rationale

21. Examples of Callout/Target pairs with difficulty level (majority voting) Target Callout Stance Rationale Difficulty the iPhone is a truly great design. I some things they get right, some things they do not. that back button is key. navigation is actually easier on android. Just because the iPhone has a huge amount of apps, doesn't they're all worth having. I feel like your poor grammar are to obvious to be self thought... disagree too. I too Some not things do Easy the `Back' button dedicated That back button is key Navigation is android Moderate much the It's more about the features and apps and Android seriously lacks on latter. - Just because the iPhone has a huge amount of apps, doesn't they're all worth having. - Difficult mean mean I feel like your comments Nexus One is too positive - Too unsure difficult/ about 21

22. Difficulty judgment (majority voting) Number of Expert Annotators per cluster Diff 5 4 3 2 1 Easy 81.0 70.8 60.9 63.6 25.0 Moderate 7.7 7.0 17.1 6.1 25.0 Difficult 5.9 5.9 7.3 9.1 12.5 Too Difficult to code 5.4 16.4 14.6 21.2 37.5 22

23. Conclusion We propose a two-tiered annotation scheme for argument annotation for online discussion forums Expert annotators detect Callout/Target pairs where crowdsourcing is employed to discover finer units like Stance/Rationale Our study also assists in detecting the text that is easy/hard to annotate Preliminary experiments to predict agreement/disagreement among ADUs 23

24. Future Work Qualitative analysis of the Callout phenomenon to process finer-grained analysis Study the different use of the ADUs on different situations Annotation on different domain (e.g. healthcare forums) and adjust our annotation scheme Predictive modeling of Stance/Rationale phenomenon 24

25. Thank you! 25

26. Example from the discussion thread User2 User3 Stance Rationale 26

27. Predicting the Agree/Disagree Relation Label Training data (143 Agree/153 Disagree) Salient Features for the experiments Baseline: rule based (`agree , `disagree ) Mutual Information (MI): MI is used to select words to represent each category LexFeat: Lexical features based on sentiment lexicons (Hu and Liu, 2004), lexical overlaps, initial words of the Callouts 10-fold CV using SVM 27

28. Predicting the Agree/Disagree Relation Label (preliminary result) Lexical features result in F1 score between 60- 70% for Agree/Disagree relations Ablation tests show initial words of the Callout is the strongest feature Rule-based system show very low recall (7%), which indicates a lot of Target-Callout relations are *implicit* Limitation lack of data (in process of annotating more data currently ) 28

29. # of Clusters for each Corpus # of EA ADUs per cluster # of Clusters Thread 5 4 3 2 1 91 52 16 11 7 5 Ipad 88 41 17 7 13 10 Layoffs 86 41 18 11 6 10 Twitter 84 44 17 14 4 5 Clusters with 5 and 4 annotators shows Callouts that are plausibly easier to identify Clusters selected by only one or two annotators are harder to identify 29

30. Target User1 Callout1 User2 Callout2 User3 30

31. Target User1 Callout1 User2 Callout2 User3 31

32. Fine-Grained Novice Annotation T T E.g., Relation Identification CO CO E.g., Agree/Disagree/ Other T T Finer-Grained Annotation CO CO E.g., Stance &Rationale 32

33. Motivation and Challenges Post1 1. Segmentation 2. Segment Classification 3. Relation Identification Post2 Post3 Post4 Argumentative Discourse Units (ADU; Peldszus and Stede, 2013) 33

34. Why we propose a two-layer annotation? A two-layer annotation schema Expert Annotation Five annotators who received extensive training for the task Primary task includes selecting discourse units from user posts (argumentative discourse units: ADU) Peldszus and Stede (2013 Novice Annotation Use of Amazon Mechanical Turk (AMT) platform to detect the nature and role of the ADUs selected by the experts 34

35. Annotation Schema for Expert Annotators Call Out A Callout is a subsequent action that selects all or some part of a prior action (i.e., Target) and comments on it in some way. Target A Target is a part of a prior action that has been called out by a subsequent action 35

36. Motivation and Challenges User generated conversational data provides a wealth of naturally generated arguments Argument mining of such online interactions, however, is still in its infancy 36

37. Detail on Corpora Four blog posts and the responses (e.g. first 100 comments) from Technorati between 2008- 2010. We selected blog postings in the general topic of technology, which contain many disputes and arguments. Together they are denoted as argumentative corpus 37

38. Motivation and Challenges (cont.) A detailed single annotation scheme seems infeasible The problem of high *cognitive load* (e.g. annotators have to read all the threads) Use of expert annotators for all tasks is costly We propose a scalable and principled two-tier scheme to annotate corpora for arguments 38

39. Annotation Schema(s) A two-layer annotation schema Expert Annotation Five annotators who received extensive training for the task Primary task includes a) segmentation, b) segment classification, and c) relation identification lecting discourse units from user posts (argumentative discourse units: ADU) Novice Annotation Use of Amazon Mechanical Turk (AMT) platform to detect the nature and role of the ADUs selected by the experts 39

40. Example from the discussion thread 40

41. A picture is worth 41

42. Motivation and Challenges 1. Segmentation 2. Segment Classification 3. Relation Identification Argument annotation includes three tasks (Peldszus and Stede, 2013) 42

43. Summary of the Annotation Schema(s) First stage of annotation Annotators: expert (trained) annotators A coarse-grained annotation scheme inspired by Pragmatic Argumentation Theory (PAT; Van Eemeren et al., 1993) Segment, label, and link Callout and Target Second stage of annotation Annotators: novice (crowd) annotators A finer-grained annotation to detect Stance and Rationale of an argument 43

44. Expert Annotation Expert Annotators Peldszus and Stede (2013) Segmentation Labeling Linking Five Expert (trained) annotators detect two types of ADUs ADU: Callout and Target Coarse-grained annotation 44

45. The Argumentative Corpus 2 1 4 3 Blogs and comments extracted from Technorati (2008-2010) 45

46. Novice Annotations: Identifying Stance and Rationale Callout Crowdsourcing Identify the task-difficulty (very difficult .very easy) Identify the text segments (Stance and Rationale) 46

47. Novice Annotations: Identifying the relation between ADUs Callout Target Crowdsourcing Relation label Number of EA ADUs per cluster 5 4 3 2 1 Agree 39.4 43.3 42.5 35.5 48.4 Disagree 56.9 31.7 32.5 25.8 19.4 Other 3.70 25.0 25.0 38.7 32.3 47

48. More on Expert Annotations Annotators were free to chose any text segment to represent an ADU Splitters Lumpers 48

49. Novice Annotation: task 1 1: Identifying the relation (agree/disagree/other) This is related to annotation of agreement/disagreement (Misra and Walker, 2013; Andreas et al., 2012) and classification of stances (Somasundaran and Wiebe, 2010) in online forums. 49

50. ADUs: Callout and Target 50

Load More ...